Artificial intelligence and unemployment dynamics: an econometric analysis in high-income economies

Abstract

Purpose

The purpose of the study is to investigate the impact of artificial intelligence (AI), machine learning (ML), and data science (DS) on unemployment rates across ten high-income economies from 2015 to 2023.

Design/methodology/approach

This study takes a unique approach by employing a dynamic panel data (DPD) model with a generalised method of moments (GMM) estimator to address potential biases. The methodology includes extensive validation through Sargan, Hansen, and Arellano-Bond tests, ensuring the robustness of the results and adding a novel perspective to the field of AI and unemployment dynamics.

Findings

The study’s findings are paramount, challenging prevailing concerns in AI, ML, and DS, demonstrating an insignificant impact on unemployment and contradicting common fears of job loss due to these technologies. The analysis also reveals a positive correlation (0.298) between larger government size and higher unemployment, suggesting bureaucratic inefficiencies that may hinder job growth. Conversely, a negative correlation (−0.201) between increased labour productivity and unemployment suggests that technological advancements can promote job creation by enhancing efficiency. These results refute the notion that technology inherently leads to job losses, positioning AI and related technologies as drivers of innovation and expansion within the labour market.

Research limitations/implications

The study’s findings suggest a promising outlook, positioning AI as a catalyst for the expansion and metamorphosis of employment rather than solely a catalyst for automation and job displacement. This insight presents a significant opportunity for AI and related technologies to improve labour markets and strategically mitigate unemployment. To harness the benefits of technological progress effectively, authorities and enterprises must carefully evaluate the balance between government spending and its impact on unemployment. This proposed strategy can potentially reinvent governmental initiatives and stimulate investment in AI, thereby bolstering economic and labour market reliability.

Originality/value

The results provide significant perspectives for policymakers and direct further investigations on the influence of AI on labour markets. The analysis results contradict the common belief of technology job loss. The study’s results are shown to be reliable by the Sargan, Hansen, and Arellano-Bond tests. It adds to the discussion on the role of AI in the future of work, proposing a detailed effect of AI on employment and promoting a strategic method for integrating AI into the labour market.

Citation

Masoud, N. (2024), "Artificial intelligence and unemployment dynamics: an econometric analysis in high-income economies", Technological Sustainability, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/TECHS-04-2024-0033